PINNs Scientist for Physics-Constrained ML

Western Digital Corporation; Singapore; Presencial; SGD 120.000 - 180.000; Jornada completa

Descripción de la vacante

WD is seeking a research scientist to develop physics-informed machine learning approaches, including PINNs, for real product problems. You will design loss functions, validate digital twins, and drive experiments with ML engineers while partnering with the storage domain expert.

Required are advanced AI/ML credentials (Master's/PhD), strong PyTorch or JAX skills, and a track record of research output. The role emphasizes prototype-to-documentation handoff and deployment readiness.

Formación

• Master's or PhD in AI/ML, Physics, Applied Mathematics or related field with strong AI/ML research focus.
• For Master's: 1-3 years in scientific ML or applied AI roles; for PhD: open—no minimum work experience; peer-reviewed publication or significant open-source contribution.
• Must have skills in PyTorch or JAX with expert-level implementation capability and PINNs methodology design.
• Area A: PINNs and physics-constrained loss architectures; Area B or C: Bayesian methods or Causal ML depending on focus; Area D: synthetic data methods.
• Demonstrated research output through publications, preprints, or significant open-source contributions; prototype-to-documentation handoff required.

Responsabilidades

• Originate and advance PINNs methodology; design physics-constrained loss function architectures and digital twin models.
• Lead research into Bayesian deep learning or causal ML depending on focus area; design acquisition functions and ensemble methods.
• Own causal ML or reliability modeling framework development for product improvements.
• Design physics-constrained generative models for synthetic data and provide training recipes for deployment.
• Interface with storage domain expert to validate physics constraints and produce validated prototypes with documentation.

Conocimientos

PyTorch/JAX expert

PINNs

Physics-constrained ML

Uncertainty Quantification

Causal ML

Digital twin modeling

Research outputs

Prototype-to-Documentation

Educación

Master's or PhD in AI/ML/Physics/Applied Math

Herramientas

Python

Scientific computing

Descripción del empleo

WD is seeking a research scientist to develop physics-informed machine learning approaches, including PINNs, for real product problems. You will design loss functions, validate digital twins, and drive experiments with ML engineers while partnering with the storage domain expert.

Required are advanced AI/ML credentials (Master's/PhD), strong PyTorch or JAX skills, and a track record of research output. The role emphasizes prototype-to-documentation handoff and deployment readiness.